The short answer
Bay Area CPA guide to cost segregation: $3M-$10M rental owners pull $300K-$500K of depreciation into year one with OBBBA bonus. San Jose tax team.
A Bay Area landlord who just closed on a $5 million 4-unit in Mountain View can pull $500,000 or more of federal tax savings into year one with a cost segregation study, instead of trickling out depreciation over 27.5 years. Land sits at $800,000, building basis at $4.2 million. Your CPA sets up straight-line depreciation under IRC §168: 27.5 years for the residential building, no salvage, midmonth convention. That gives you about $152,000 of depreciation per year for the next 27 and a half years. Useful, but boring.
A cost segregation study, run by a qualified engineer, can break that same $4.2 million basis into five-year, fifteen-year, and 27.5-year buckets. Combined with the 100% bonus depreciation that came back under the One Big Beautiful Bill Act (OBBBA, Public Law 119-21, signed July 4, 2025), your year-one deduction on the same property jumps to roughly $1.1 million. The study costs $5,000 to $15,000. The tax savings can clear $500,000.
OBBBA timing note: the 100% bonus depreciation revival under OBBBA applies to property placed in service after January 19, 2025. Property placed in service on or before that date still follows the prior-law phase-down (40% in 2025, 20% in 2026, zero thereafter). Confirm placed-in-service date before running the year-one numbers.
For a Bay Area landlord buying anything in the $3M to $10M range in San Jose, Mountain View, Sunnyvale, Cupertino, or Palo Alto, the math is almost always worth running. This guide walks through what cost segregation actually is, how the depreciation classes split, how OBBBA changed the bonus math, who can actually use the losses (this is where most Bay Area earners get tripped), and when to skip the study entirely.
Under §168, residential rental property is depreciated straight-line over 27.5 years. Commercial property (offices, retail, industrial) gets 39 years. Both are §1250 property: real property, structural components, the building itself.
On a $4.2 million Mountain View residential basis, that's $152,727 per year of depreciation. On the same $4.2 million treated as commercial, you'd get $107,692 per year. Either way, the deduction trickles in slowly over decades. If you sell in year seven, you only ever claimed about $1 million of the $4.2 million in depreciation, and the rest dies with the building (or transfers to the buyer via their stepped basis).
Cost segregation attacks this slow trickle by reclassifying portions of the building into shorter-life §1245 personal property and §1250 land improvements. Carpet, decorative lighting, appliances, dedicated electrical for kitchens, parking lot asphalt, landscaping, fences: all of it can come out of the 27.5/39-year bucket and into 5-year, 7-year, or 15-year buckets. When you front-load depreciation into those shorter buckets and then apply bonus depreciation on top, the year-one deduction explodes.
The IRS publishes the framework for this analysis in the Cost Segregation Audit Techniques Guide. The guide describes the engineering-based methodology auditors expect: a qualified engineer or specialist physically (or virtually, for newer construction with full drawings) inspects the property, reviews blueprints and cost ledgers, and assigns each building component to its correct asset class under MACRS.
The buckets the study produces typically look like this:
A well-run study produces a defensible report: photos, blueprints, cost ledger, asset-class assignments tied to specific Treasury rulings and court cases. That report is what you hand to the IRS if anyone asks. Cheap studies (under $3,000) often skip the engineering work and rely on rules of thumb. Those don't hold up in audit.
The other half of the math is §168(k) bonus depreciation. The 2017 Tax Cuts and Jobs Act gave us 100% bonus depreciation on qualifying property (anything with a recovery period of 20 years or less) placed in service from late 2017 through 2022. Starting 2023, that bonus phased down: 80% in 2023, 60% in 2024, 40% in 2025, 20% in 2026, zero in 2027.
OBBBA (Pub. L. 119-21) reversed that schedule. For property placed in service after January 19, 2025, bonus depreciation snaps back to 100% and is permanent. Property placed in service on or before January 19, 2025 still follows the pre-OBBBA phase-down (40% bonus in 2025).
What this means for cost seg under OBBBA: every dollar your study reclassifies into 5-year, 7-year, or 15-year property is eligible for 100% bonus depreciation in year one. The 27.5-year and 39-year remainder still depreciates straight-line over its full life. The more aggressive (and defensible) the study, the bigger the year-one deduction. For property placed in service before January 19, 2025 (still on the 40% bonus phase-down), that same reclassified personal property gets 40% bonus plus accelerated MACRS on the remainder, which still beats straight-line 27.5/39-year but reduces the year-one number by roughly 50% to 60% versus the 100% bonus scenario.
Take a Sunnyvale 8-unit purchased June 1, 2025 for $5 million. The tax assessor and appraisal allocate $800,000 to land, leaving $4.2 million in depreciable building basis. The buyer files Schedule E as a passive investor at first, then qualifies for real estate professional status in 2026.
A qualified engineer runs a cost seg study for $12,000 and produces this breakdown:
| Asset Class | Recovery Period | Reclassified Basis | Year-1 Deduction (100% bonus) |
|---|---|---|---|
| 5-year property (appliances, carpet, dedicated electrical) | 5 years | $850,000 | $850,000 |
| 15-year land improvements (parking, landscaping) | 15 years | $250,000 | $250,000 |
| 27.5-year building shell | 27.5 years | $3,100,000 | ~$56,400 (straight-line, partial year) |
| Total year-1 depreciation | $4,200,000 | ~$1,156,400 |
Without cost seg, the same year-one deduction would have been roughly $89,000 (partial-year straight-line on the full $4.2M basis). The study generated about $1,067,000 of additional year-one depreciation under the 100% bonus scenario.
Pre-OBBBA (placed-in-service on or before January 19, 2025) version of the same numbers. Under the prior phase-down with 2025 bonus at 40%, the same study produces a smaller year-one number: the $850K of 5-year property gets $340K bonus (40%) plus first-year MACRS on the remaining $510K (roughly $102K) = about $442K year-one deduction on the 5-year bucket. The $250K of 15-year improvements gets $100K bonus plus a small MACRS slice. Total year-one accelerated deduction roughly $585K instead of $1,100K. Still substantially better than the $89K straight-line baseline, but the cost-seg ROI is roughly halved.
At a 37% federal marginal rate, 9.3% California marginal rate, and 3.8% Net Investment Income Tax (where applicable), the combined marginal rate clears 50%. On $1,067,000 of accelerated deductions under OBBBA's 100% bonus, the federal-plus-state-plus-NIIT tax shield is roughly $530,000 in year one alone. Under the 40% pre-OBBBA scenario the shield drops to roughly $250,000. Study fee: $12,000 in either case.
Important NIIT note for the worked example. NIIT only applies to passive-investor scenarios. A landlord who qualifies for real estate professional status (REPS) and treats the rental as a non-passive trade or business does NOT pay 3.8% NIIT on the net rental income (or recapture later at sale). The 50% combined marginal rate above is the passive-investor ceiling; REPS landlords typically land closer to 46% combined (37% federal + 9.3% CA, no NIIT) on active rental income.
Running cost seg numbers for your own Bay Area property? We model the OBBBA-vs-prior-law deltas (driven by placed-in-service date), REPS feasibility, and recapture exposure before any client commissions a study. Schedule a complimentary consultation.
This is where most Bay Area W-2 earners get stopped. Rental real estate is, by default, a passive activity under IRC §469. Passive losses can only offset passive income. They cannot offset your W-2 salary, your RSU income, or your interest and dividends.
There are two ways out:
If you actively participate in your rental (meaning you make management decisions, even if you hire out the day-to-day), you can deduct up to $25,000 of passive losses against ordinary income. But the allowance phases out completely between $100,000 and $150,000 of modified AGI. A Bay Area software engineer earning $250,000 is already above the phaseout, and a married couple with two tech incomes is so far above it the §469(i) door is closed.
Bottom line: for almost every Bay Area earner thinking about cost seg on their own property, §469(i) is irrelevant.
Real estate professional status (REPS) is the path that works for high earners. To qualify, the taxpayer (or one spouse on a joint return) must:
The second test is the killer for tech workers. If you have a full-time W-2 job, you cannot meet the 50% test. There is no way to spend more time on rentals than on your day job while also having a day job. This is why REPS is realistically available to: (a) one spouse who has left their W-2 and runs the rental portfolio full-time, (b) full-time real estate developers and brokers, and (c) retirees who self-manage substantial portfolios.
The classic Bay Area cost seg structure is one spouse keeping the W-2 income, the other spouse claiming REPS, and the accelerated rental losses flowing against the working spouse's salary on a joint return. Material participation hours need contemporaneous logs: time, date, activity, property. If the IRS audits and the log is reconstructed after the fact, REPS gets denied and the losses get suspended.
Suspended passive losses carry forward indefinitely. They release when (a) you have offsetting passive income, (b) you qualify for REPS in a later year, or (c) you sell the property in a fully taxable disposition. So even passive investors get to use the cost seg eventually, just not in year one.
Accelerated depreciation is not a free lunch. When you sell, the IRS recaptures the depreciation you took. The two recapture regimes work very differently:
For a cost-seg-heavy property, a meaningful chunk of your eventual gain on sale becomes ordinary-rate recapture instead of long-term capital gain. The net benefit is the time value of money: deducting at 50% combined marginal rate today, paying 40% combined marginal rate at sale ten years from now. As long as you actually have a marginal rate to shield against, the present-value math almost always works.
Two ways to soften the recapture hit: a 1031 exchange at sale (defers the gain into a replacement property), or holding until death (basis steps up, depreciation recapture eliminated). Both are real planning levers and both should be modeled before you commission the study.
You don't have to do the cost seg in the year you buy. The IRS allows a "look-back" study on properties you've held for years, and the missed depreciation comes through as a §481(a) adjustment in the year of the change. The mechanism is Form 3115, Application for Change in Accounting Method.
Example: you bought a 6-unit in Cupertino in 2021 for $4 million, allocated $3.4 million to building. You've taken about $620,000 of straight-line depreciation through 2025. A cost seg run in 2026 would have produced $850,000 in additional accelerated depreciation if you'd done it from day one. You file Form 3115 with your 2026 return and pick up that $850,000 of "missed" depreciation as a single §481(a) adjustment on the 2026 return. No amended returns needed for prior years.
The look-back is one of the best cost seg use cases for Bay Area investors who never asked the question on properties they already own. We see this constantly with clients who inherited rentals from older relatives, or who bought before they ever had a CPA who knew to mention it.
The most common mistake we see at intake on cost seg files: a Bay Area buyer commissions a $3,000 "calculator" study from an online provider, takes the year-one deduction, and the IRS opens the file three years later. With no engineering report, no photos, no cost ledger tied to specific Treasury rulings, the entire reclassification gets disallowed and the depreciation flows back as ordinary-rate recapture in the audit year. Cost of the mistake: $150,000 to $400,000 of recaptured tax plus penalties, against $3,000 saved on study fee.
The second pattern: filers who run cost seg without confirming REPS for the household. The accelerated losses suspend as passive activity losses and provide zero current-year benefit. The study was still worth running (the losses release at sale or in a later REPS year) but the client expected a $250K refund check that never arrives. TurboTax does not flag the §469 trap. DIY filing software handles the basic mechanics; it does not flag REPS feasibility, the California depreciation delta, or the §481(a) look-back election. Those are the items where engagement pays for itself many times over.
Cost seg isn't always worth it. The clear cases to skip:
No. Cost seg is only available on property held for investment or used in a trade or business. Your primary residence is personal-use property and gets no depreciation at all. If you convert a former residence into a rental, you can do cost seg from the conversion date forward, using the lower of basis or fair market value on the conversion date.
You can still run a look-back cost segregation study and recapture the missed depreciation through a §481(a) adjustment on Form 3115 in the year of the method change. No amended returns required. This works on properties held for any length of time, as long as you still own them.
No. California does not conform to §168(k) bonus depreciation, and never has. For California purposes, all your reclassified 5/7/15-year property depreciates under California's modified MACRS (similar to federal pre-bonus rules). The cost seg study still benefits you for federal purposes, but California will pick up its share of tax on a much slower schedule. We track this with a separate California depreciation schedule on every cost-seg client.
For a Bay Area rental in the $3M-$10M range, a qualified engineering study typically runs $5,000 to $15,000 depending on property complexity, age, and whether construction documents are available. Avoid anyone offering studies under $3,000: those are usually rule-of-thumb estimates without engineering support and will not survive audit.
For Bay Area earners above the §469(i) AGI phaseout (which is essentially everyone with a tech income), yes, REPS is the only way to use the losses against ordinary income in the year you take them. Without REPS, the accelerated losses are suspended as passive activity losses and carry forward until you have passive income or sell the property. The losses are not lost, but the year-one tax benefit doesn't materialize.
The engineering report your study provider produces is the defense. A study built around the IRS Cost Segregation Audit Techniques Guide methodology (with photos, cost ledger, asset class rationale tied to specific rulings) holds up. Cheap "calculator" studies do not. This is why provider selection matters as much as the study itself.
We work with several engineering firms that produce defensible cost seg studies in the Bay Area, and we run the full tax model before any client commits to one. For a typical engagement, we:
If you bought a Bay Area rental in the last five years, or you're under contract on one now, the cost seg conversation is worth having before you file the year of acquisition. The OBBBA 2025 100% bonus depreciation revival makes the math more aggressive than it's been since 2022. Our tax planning team handles this for real estate investors across the Peninsula and South Bay. Many of our tech employee clients are also landlords, and the cost seg conversation often comes up alongside RSU planning and AMT.
Schedule a complimentary consultation and we'll model the depreciation math, recapture exposure, and whether REPS makes sense for your household before you commit to a study.
Cost seg plus 100% bonus depreciation can pull six or seven figures of deduction into year one. We run the math before you commission anything.